Instructions to use WindyWord/translate-es-zai with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use WindyWord/translate-es-zai with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "translation" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # 'pip install "transformers<5.0.0' from transformers import pipeline pipe = pipeline("translation", model="WindyWord/translate-es-zai")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("WindyWord/translate-es-zai", device_map="auto") - Notebooks
- Google Colab
- Kaggle
WindyWord.ai Translation β Spanish β Isthmus Zapotec
Part of the Windstorm Labs open model catalogue. Scores, licence and attribution for every model: https://windytranslate.com/models/translate-es-zai
Canonical copy: https://huggingface.co/WindyTranslate/translate-es-zai
Licence under review: trained on the eBible corpus, which includes translations under non-commercial / no-derivatives terms. Not for commercial use until this notice is removed.
This applies to the WindyScripture variants in this repository (
herm0-scripture/,scripture-ct2-int8/). The other variants are derived from the upstream model under its licence, shown in the Provenance & License section below.
Translates Spanish β Isthmus Zapotec.
Quality
No quality score is published in this repository. Current screening scores, where measured, are on the catalogue page linked above.
Available Variants
Deployment formats in this repository (subfolders):
| Variant | Description |
|---|---|
lora/ |
WindyStandard β production baseline. Transformers format for GPU inference. |
lora-ct2-int8/ |
WindyStandard Β· CPU INT8 β CTranslate2 INT8 quantization of WindyStandard for CPU inference. |
herm0-scripture/ |
WindyScripture β verse-aligned fine-tune on the eBible parallel corpus. Specialised for biblical text; not recommended for general translation. Licence under review β see notice above. |
scripture-ct2-int8/ |
WindyScripture Β· CPU INT8 β CTranslate2 INT8 quantization of WindyScripture. Licence under review β see notice above. |
Quick usage
Transformers (PyTorch):
from transformers import MarianMTModel, MarianTokenizer
tokenizer = MarianTokenizer.from_pretrained("WindyWord/translate-es-zai", subfolder="lora")
model = MarianMTModel.from_pretrained("WindyWord/translate-es-zai", subfolder="lora")
CTranslate2 (fast CPU inference):
import ctranslate2
translator = ctranslate2.Translator("path/to/translate-es-zai/lora-ct2-int8")
Apps
The Windy Word apps are built on this model family.
Provenance & License
Weights derived from Helsinki-NLP/opus-mt-es-zai (OPUS-MT, Helsinki-NLP, University of Helsinki), licensed Apache-2.0. Windy variants are released under the same licence.
Exception: the WindyScripture variants (herm0-scripture/, scripture-ct2-int8/) are under licence review β see the notice at the top of this card. Not for commercial use until that notice is removed.
Model tree for WindyWord/translate-es-zai
Base model
Helsinki-NLP/opus-mt-es-zai